中文摘要：

针对均匀线列阵，在宽带混合信号（不相关和相干信号共存）情形下，提出了一种DOA快速估计新算法。利用阵列协方差矩阵的Hermitian性，通过酉变换将各频点的复数据矩阵映射为实矩阵，通过实值化的TOFS法先直接估计出宽带非相关信号的DOA；然后利用空间差分技术，在各个频点上得到只含相干信号的数据协方差矩阵；通过Toeplitz矩阵重构，在不降低阵列孔径的条件下，可实现相干信号的解相干，再利用实值TOFS法可得到相干信号的DOA。由于算法是并行分别对不相关和相干信号进行DOA估计，在信源过载（信号数大于阵元数）的情形下，算法依然有效，同时由于实值化，算法的计算复杂度较小。仿真结果验证了算法的有效性。

英文摘要：

This paper proposed a new algorithm for DOA （ direction of arrival） estimation under the coexistence of both uncor- related and coherent wideband signals with uniform linear array（ULA）. Firstly, according to the Hermitian feature of array ma- trix, through unitary transform, complex matrix could convert into real matrix. And then, through unitary TOFS algorithm, it obtained DOA estimates of uncorrelated signals. Secondly, using differencing technique, it could acquire coherent signals data matrix, and through Toeplitz technique of matrix, coherent wideband signals could convert into uneorrelated signals without any division for the array. After that, using unitary TOFS algorithm, it obtained DOA estimates of coherent signals. The algorithm can resolve more wideband signals than the array elements because of the parallel of the algorithm, and at the same time, the computational complexity of the algorithm is low for the feature of real value. Simulation results demonstrate its effectiveness.